verifier keeps track of register state spilled to stack.
registers are 8-byte wide and always aligned, so instead of tracking them
in every byte-sized stack slot, use MAX_BPF_STACK / 8 array to track
spilled register state.
Though verifier runs in user context and its state freed immediately
after verification, it makes sense to reduce its memory usage.
This optimization reduces sizeof(struct verifier_state)
from 12464 to 1712 on 64-bit and from 6232 to 1112 on 32-bit.

Note, this patch doesn't change existing limits, which are there to bound
time and memory during verification: 4k total number of insns in a program,
1k number of jumps (states to visit) and 32k number of processed insn
(since an insn may be visited multiple times). Theoretical worst case memory
during verification is 1712 * 1k = 17Mbyte. Out-of-memory situation triggers
cleanup and rejects the program.
Suggested-by: NAndy Lutomirski <luto@amacapital.net>
Signed-off-by: NAlexei Starovoitov <ast@plumgrid.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Guenter Roeck says:

====================
net: dsa: Fixes and enhancements

Patch 01/15 addresses a bug indicated by an an annoying and unhelpful
log message.

Patches 02/15 and 03/15 are minor enhancements, adding support for
known switch revisions.

Patches 04/15 and 05/15 add support for MV88E6352 and MV88E6176.

Patch 06/15 adds support for hardware monitoring, specifically for
reporting the chip temperature, to the dsa subsystem.

Patches 07/15 and 08/15 implement hardware monitoring for MV88E6352,
MV88E6176, MV88E6123, MV88E6161, and MV88E6165.

Patch 09/15 and 10/15 add support for EEPROM access to the DSA subsystem.

Patch 11/15 implements EEPROM access for MV88E6352 and MV88E6176.

Patch 12/15 adds support for reading switch registers to the DSA
subsystem.

Patches 13/15 amd 14/15 implement support for reading switch registers
to the drivers for MV88E6352, MV88E6176, MV88E6123, MV88E6161, and MV88E6165.

Patch 15/15 adds support for reading additional RMON registers to the drivers
for  MV88E6352, MV88E6176, MV88E6123, MV88E6161, and MV88E6165.

The series was tested on top of v3.18-rc2 in an x86 system with MV88E6352.
Testing in systems with 88E6131, 88E6060 and MV88E6165 was done earlier
(I don't have access to those systems right now). The series was also build
tested using my build system at http://server.roeck-us.net:8010/builders.
Look into the 'dsa' column for build results.

The series merges cleanly into net-next as of today (10/29).

v3:
- Fix bug in eeprom patches seen if devicetree is enabled:
  eeprom-length property is attached to switch devicetree node,
  not to dsa node, and there was a compile error.
v2:
- Made reporting chip temperatures through the hwmon subsystem optional
  with new Kconfig option
- Changed the hwmon chip name to <network device name>_dsa<index>
- Made EEPROM presence and size configurable through platform and devicetree
  data
- Various minor changes and fixes (see individual patches for details)
====================
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Display sw_in_discards, sw_in_filtered, and sw_out_filtered for chips
supported by mv88e6123_61_65 and mv88e6352 drivers.

The variables are provided in port registers, not the normal status registers.
Mark by adding 0x100 to the register offset and add special handling code
to mv88e6xxx_get_ethtool_stats.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Report switch register values to ethtool.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The infrastructure can now report switch registers to ethtool.
Add support for it to the mv88e6123_61_65 driver.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add support for reading switch registers with 'ethtool -d'.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

MV88E6352 supports read and write access to its configuration eeprom.

There is no means to detect if an EEPROM is connected to the switch.
Also, the switch supports EEPROMs with different sizes, but can not detect
or report the type or size of connected EEPROMs. Therefore, do not implement
the get_eeprom_len callback but depend on platform or devicetree data to
provide information about EEPROM presence and size.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The dsa core now supports reading from and writing to a switch EEPROM
if connected. Describe optional devicetree property indicating that
an EEPROM is present and its size.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

On some chips it is possible to access the switch eeprom.
Add infrastructure support for it.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

MV88E6123 and compatible chips support reading the chip temperature
from PHY register 6:26.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

MV88E6352 supports reading the chip temperature from two PHY registers,
6:26 and 6:27. Report it using the more accurate register 6:27.
Also report temperature limit and alarm.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Some switches provide chip temperature data.
Add support for reporting it through the hwmon subsystem.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

MV88E6176 is mostly compatible to MV88E6352 and is documented
in the same functional specification. Add support for it.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Marvell 88E6352 is mostly compatible to MV88E6123/61/65,
but requires indirect phy access. Also, its configuration
registers are a bit different.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Report known silicon revisions when probing Marvell 88E6131 switches.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Report known silicon revisions when probing Marvell 88E6060 switches.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Setting skb->protocol to a private protocol type may result in warning
messages such as
	e1000e 0000:00:19.0 em1: checksum_partial proto=dada!

This happens if the L3 protocol is IP or IPv6 and skb->ip_summed is set
to CHECKSUM_PARTIAL. Looking through the code, it appears that changing
skb->protocol for transmitted packets is not necessary and may actually
be harmful. For example, it prevents purposely unmodified (from a DSA
perspective) network drivers from properly setting up their transmit
checksum offload pointers since they inspect skb->protocol to set up the
IPv4 header or IPv6 header pointers. So don't unnecessarily change the
protocol field.
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In neigh_parms_release() we loop over all entries to find the entry given in
argument and being able to remove it from the list. By using a double linked
list, we can avoid this loop.

Here are some numbers with 30 000 dummy interfaces configured:

Before the patch:
$ time rmmod dummy
real	2m0.118s
user	0m0.000s
sys	1m50.048s

After the patch:
$ time rmmod dummy
real	1m9.970s
user	0m0.000s
sys	0m47.976s
Suggested-by: NThierry Herbelot <thierry.herbelot@6wind.com>
Signed-off-by: NNicolas Dichtel <nicolas.dichtel@6wind.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

napi_schedule() can be called from any context and has to mask hard
irqs.

Add a variant that can only be called from hard interrupts handlers
or when irqs are already masked.

Many NIC drivers can use it from their hard IRQ handler instead of
generic variant.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

David Vrabel says:

====================
xen-netback: minor cleanups

Two minor xen-netback cleanups originally from Zoltan.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This flag is unnecessary, it came from some old code.
Suggested-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NZoltan Kiss <zoltan.kiss@linaro.org>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>
Acked-by: NWei Liu <wei.liu2@citrix.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Otherwise the interrupt handler still calls napi_complete. Although it
won't schedule NAPI again as either NAPI_STATE_DISABLE or
NAPI_STATE_SCHED is set, it is just unnecessary, and it makes more
sense to do this way.
Signed-off-by: NZoltan Kiss <zoltan.kiss@linaro.org>
Signed-off-by: NDavid Vrabel <david.vrabel@citrix.com>
Acked-by: NWei Liu <wei.liu2@citrix.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add a sysctl that causes an interface's optimistic addresses
to be considered equivalent to other non-deprecated addresses
for source address selection purposes.  Preferred addresses
will still take precedence over optimistic addresses, subject
to other ranking in the source address selection algorithm.

This is useful where different interfaces are connected to
different networks from different ISPs (e.g., a cell network
and a home wifi network).

The current behaviour complies with RFC 3484/6724, and it
makes sense if the host has only one interface, or has
multiple interfaces on the same network (same or cooperating
administrative domain(s), but not in the multiple distinct
networks case.

For example, if a mobile device has an IPv6 address on an LTE
network and then connects to IPv6-enabled wifi, while the wifi
IPv6 address is undergoing DAD, IPv6 connections will try use
the wifi default route with the LTE IPv6 address, and will get
stuck until they time out.

Also, because optimistic nodes can receive frames, issue
an RTM_NEWADDR as soon as DAD starts (with the IFA_F_OPTIMSTIC
flag appropriately set).  A second RTM_NEWADDR is sent if DAD
completes (the address flags have changed), otherwise an
RTM_DELADDR is sent.

Also: add an entry in ip-sysctl.txt for optimistic_dad.
Signed-off-by: NErik Kline <ek@google.com>
Acked-by: NLorenzo Colitti <lorenzo@google.com>
Acked-by: NHannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Hayes Wang says:

====================
r8152: support nway_reset

Fix the CHECK from checkpatch.pl and support nway_reset.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Support the nway_reset() function for ethtool.
Signed-off-by: NHayes Wang <hayeswang@realtek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Replace tx_underun with tx_underrun for checkpatch.pl.
Signed-off-by: NHayes Wang <hayeswang@realtek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While testing upcoming Yaogong patch (converting out of order queue
into an RB tree), I hit the max reordering level of linux TCP stack.

Reordering level was limited to 127 for no good reason, and some
network setups [1] can easily reach this limit and get limited
throughput.

Allow a new max limit of 300, and add a sysctl to allow admins to even
allow bigger (or lower) values if needed.

[1] Aggregation of links, per packet load balancing, fabrics not doing
 deep packet inspections, alternative TCP congestion modules...
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Yaogong Wang <wygivan@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Several warnings and errors of coding style rules corrected.
Compile tested.
Signed-off-by: NRoberto Medina <robertoxmed@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch generalizes commit d6a4a104 ("tcp: GSO should be TSQ
friendly") to protocols using skb_set_owner_w()

TCP uses its own destructor (tcp_wfree) and needs a more complex scheme
as explained in commit 6ff50cd5 ("tcp: gso: do not generate out of
order packets")

This allows UDP sockets using UFO to get proper backpressure,
thus avoiding qdisc drops and excessive cpu usage.

Here are performance test results (macvlan on vlan):

- Before
# netperf -t UDP_STREAM ...
Socket  Message  Elapsed      Messages
Size    Size     Time         Okay Errors   Throughput
bytes   bytes    secs            #      #   10^6bits/sec

212992   65507   60.00      144096 1224195    1258.56
212992           60.00          51              0.45

Average:        CPU     %user     %nice   %system   %iowait    %steal     %idle
Average:        all      0.23      0.00     25.26      0.08      0.00     74.43

- After
# netperf -t UDP_STREAM ...
Socket  Message  Elapsed      Messages
Size    Size     Time         Okay Errors   Throughput
bytes   bytes    secs            #      #   10^6bits/sec

212992   65507   60.00      109593      0     957.20
212992           60.00      109593            957.20

Average:        CPU     %user     %nice   %system   %iowait    %steal     %idle
Average:        all      0.18      0.00      8.38      0.02      0.00     91.43

[edumazet] Rewrote patch and changelog.
Signed-off-by: NToshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The hardware can automatically generate pause frames when the number
of free buffers drops under a certain threshold, but in order to do this,
the address of the last free buffer needs to be written to a specific
register for each RX queue.

This has to be done in 'gfar_clean_rx_ring' which is called for each
RX queue. In order not to impact performance, by adding a register write
for each incoming packet, this operation is done only when the PAUSE frame
transmission is enabled.

Whenever the link is readjusted, this capability is turned on or off.
Signed-off-by: NMatei Pavaluca <matei.pavaluca@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Local flow control options needed in order to resolve the negotiation
are incorrectly calculated.

Previously 'mii_advertise_flowctrl' was called to determine the local advertising
options, but these were determined based on FLOW_CTRL_RX/TX flags which are
never set through ethtool.
The patch simply translates from ethtool flow options to mii flow options.
Signed-off-by: NPavaluca Matei <matei.pavaluca@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The phy device supports 802.3x flow control, but the specific flags are not set
in the phy initialisation code. Flow control flags need to be added to the
supported capabilities of the phydev by the driver.

This is needed in order for ethtool to work ('ethtool -A' code checks for these
flags)
Signed-off-by: NPavaluca Matei <matei.pavaluca@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ERROR: "lockdep_ovsl_is_held" [net/openvswitch/vport-gre.ko] undefined!
Reported-by: NAlexei Starovoitov <alexei.starovoitov@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Amir Vadai says:

====================
Mellanox ethernet driver update Oct-27-2014

This patchset introduces some small bug fixes, support in get/set of
vlan offload and get/set/capabilities of the link.

First 7 patches by Saeed, add support in setting/getting link speed and getting
cable capabilities.
Next 2 patches also by Saeed, enable the user to turn rx/tx vlan offloading on
and off.
Jenni fixed a bug in error flow during device initalization.
Ido and Jack fixed some code duplication and errors discovered by static checker.
last patch by me is a fix to make ethtool report the actual rings used by
indirection QP.

Patches were applied and tested against commit 61ed53de ("Merge tag 'ntb-3.18'
of git://github.com/jonmason/ntb")
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Hardware requires the number of rings in indirection table to be a power
of 2. When setting number of channels to a non power of 2 number,
indirection table is using only the closest power of 2 rings.
Report this number in 'ethtool -x' and not the total number of rx rings.
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NEugenia Emantayev <eugenia@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Upon failures, destroy_netdev is called, and spinlocks/works must be
initialized before calling it. Otherwise kernel panic may occur.
Signed-off-by: NEugenia Emantayev <eugenia@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is instead of calling the actual implementation of
napi_synchronize, for better encapsulation.
Signed-off-by: NIdo Shamay <idos@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

clang flagged the following. All are actually cosmetic cleanups, not really bugs:

drivers/net/ethernet/mellanox/mlx4/en_main.c:233:3: warning: Value stored to 'err' is never read
                err = -ENOMEM;
                ^     ~~~~~~~
drivers/net/ethernet/mellanox/mlx4/en_main.c:293:3: warning: Value stored to 'err' is never read
                err = -ENOMEM;

drivers/net/ethernet/mellanox/mlx4/en_netdev.c:648:16: warning: Assigned value is garbage or undefined
        entry->reg_id = reg_id;
                      ^ ~~~~~~
drivers/net/ethernet/mellanox/mlx4/en_netdev.c:659:2: warning: Function call argument is an uninitialized value
        mlx4_en_uc_steer_release(priv, priv->dev->dev_addr, *qpn, reg_id);
(NOTE: reg_id is only used in the device-managed flow steering path, in which is it always initialized.
 This is not a bug. Cleanup here is therefore cosmetic only).

drivers/net/ethernet/mellanox/mlx4/en_rx.c:122:3: warning: Value stored to 'frag_info' is never read
                frag_info = &priv->frag_info[i];
                ^           ~~~~~~~~~~~~~~~~~~~
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Move mlx4_en_reset_config to en_netdev.c as it now serves more general purpose.
Add support for turning OFF/ON the rx/tx vlan offlad.
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Rename mlx4_en_timestamp_config to mlx4_en_reset_config and extend it to support
choosing RX vlan offload configuration.
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>